Demand Forecast Device Using Segmented Visit and Ratio Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing demand prediction technologies cannot accurately predict the number of sales of a commodity in a store, as they fail to distinguish between the influence of the number of visits and the popularity of the commodity.
Innovation Solution
A demand prediction device that includes an acquisition unit for gathering factor information, a visit prediction unit using a number-of-visits prediction model, a ratio prediction unit using a ratio prediction model, and a sales prediction unit to forecast the number of sales by combining the predicted number of visits and sales ratio, considering environmental and feature information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single demand prediction model is used to predict the number of visits to a store, then the prediction process is simple, but the prediction accuracy of commodity sales cannot be achieved because it fails to distinguish between the influence of visits and commodity popularity
Solution Approach 1:
The patent divides the demand prediction into two separate prediction models: a first prediction model for predicting the number of visits to the store, and a second prediction model for predicting the sales ratio of each commodity. This segmentation allows each model to focus on specific aspects (visit behavior and purchasing behavior) thereby improving overall prediction accuracy while maintaining manageable complexity through modular design
Solution Approach 2:
The patent introduces an intermediary variable (sales ratio) that bridges the gap between store visits and commodity sales. The first prediction model outputs predicted visits, which then serve as input to the second prediction model that calculates the sales ratio, and finally these are combined to predict actual sales. This intermediary approach enables accurate commodity sales prediction by separating the influences of visit volume and purchasing behavior
2Adaptability or versatility
If commodity features are kept static in the prediction model, then the model is simpler to manage, but it cannot account for changes in commodity popularity over time
Solution Approach 1:
The patent implements dynamic commodity features that can change over time. The commodity feature information includes attributes such as commodity category, price, and promotional status that are updated periodically or in response to changing conditions. This dynamic approach allows the prediction model to adapt to changes in commodity popularity and characteristics, improving the accuracy of sales predictions for commodities with changing features
Solution Approach 2:
The system incorporates feedback mechanisms where actual sales data and changing commodity features are fed back into the prediction models to continuously improve accuracy. The second prediction model learns from historical data about commodity popularity changes and adjusts its predictions accordingly, enabling the system to adapt to temporal variations in commodity demand patterns
Data Source
AI summary
A demand prediction device predicts the number of visits based on a number-of-visits prediction model used to predict the number of visits and factor information regarding a factor influencing the number of sales, predicts a ratio based on a ratio prediction model used to predict a ratio of the number of sales to the number of visits, and the factor information and predicts the number of sales of the commodity based on the predicted number of visits and the predicted ratio.